Issue Encountered While Setting Up DMFold

[jlspzw]

Hi Xin-Dong,

Could you also send me your local version model (ideally also including the MSAs), I will see the difference.

BTW, from the model, I can see the terminal helix seems without contact with other parts, so it is possible that the orientation between the terminal helix with other parts has a different results indifferent runs

[Xin-Dong Xu]

Hi Wei,

I am writing to confirm whether you have received the files I shared with you via email.

I attached the working directory of my DMFold run, which includes the model and multiple sequence alignment files. Kindly let me know if you have successfully received the files.

Thank you once again for your prompt attention to this matter. I appreciate your assistance and look forward to your response.

Best regards,
Xin-Dong

[jlspzw]

Hi Xin-Dong,

Please see the structure comparison results of your local model (yellow+green) and server model (purple)

First, based on my checking, the pLDDT score of those two models is around 0.75, which is >0.7 indicating they are correct backbone folds but may not be perfect as pLDDT ranges from 0-1.

Second, the model contains three main parts, loop (left, red circle) region, beta core (center, blue circle), and alpha helix (right, red circle), you can see except loop region, the beta core and helix region matched well (RMSD of the core region are around 0.15), but the orientation between helix and core region is slightly different, that is why if you compare the global structure they achieve 2A RMSD if you check the model (or contact map in web server) and MSA, you can see there are almost no contact between core region and helix, which means the orientation is difficult to predict. Thus the orientation of the local model and server model is quite different. In addition, the loop region also contributes a lot to the RMSD, and as we know loop is also difficult to predict (see residue-level pLDDT).

In my opinion, the two models have almost identical local regions, but the loop and orientation are slightly different leading to the RMSD seeming big.

Let me know if you have any questions.

Best Regards
Wei Zheng

compare.PNG (528.22 KiB) Viewed 6386 times

[Xin-Dong Xu]

Hi Wei,

I am very grateful for your insightful response. It turns out that the instability lies in the structure of this protein, and I mistakenly thought there was an issue with my installation. Thank you for all your help and I look forward to learning from you again. Wishing you all the best.

Best regards,
Xin-Dong